AVR-Chat

In my previous attempt (and research) I did exactly what you 
recommend and it worked good (big 105 400V caps). However, in the 
final package of the device, I surprisingly found the 'traces' of the 
same old symptom. By the way, my 1/2 watt carbon resistors make smoke 
and burn! Anyway, I still have them parallel (and as close as 
possible) to contactors. Not enough, apparently. 

Today I found a part (not all though) of problem comes from 5V 
relays. Maybe a solid state relay can help. Suggestions?

--- In AVR-Chat@yahoogroups.com, maria mastik <maman_barghi2000@...> 
wrote:
>
> Hi
>   Simply connect a 1uF 600V (or 400V) ceramic capasitor in parallel 
with the input chock of each contactor (if you have more than one); 
this will eleminate the problem compleately.
>    
>   anothe way is to connect a 670 Ohm resistor with the capasitor in 
series and then connect thease in parallel to chock of contactor.
>    
>   the masures which I mentiond  above (I mean 670 OHM and 1 uF) are 
through industerial experiance.
>    
>   Regards
>   Mashaghi.
> 
> behrooz_hariri <behrooz_hariri@...> wrote:
>           
> Problem:
> 
> Industrial contactors make such a high noise (surge) that 
occasionally
> cause these symptoms on AVR MCU :
> 
> 1- LCD blink, unrecognized characters
> 
> 2- Flash rom bitmap change
> 
> 3- EEPROM change or erasure
> 
> Explanation:
> 
> My atmega16, controls 3-phase contactors
> <http://en.wikipedia.org/wiki/Contactor> through 5V relays which are
> using ULN2803. As chokes, I wound about 10 turn of wire over a 8mm
> diameter ferrite rod taken from AM radio antenna. So, MCU commands 
to
> ULN, it activates 5V relays, and after the rod choke, contactors (in
> another metal case) are trigged. I used the same (one) rod for all
> input/output lines in controller box. This made the situation much
> better, yet I get the same problem after 100-200 cycle of contactor 
work
> (for exaple after a 5 hours of work).
> 
> My Understaing:
> 
> I guess when 220V AC sine wave gets it peack, and contactors is
> released at the same moment, the maximum spark is created and it
> reflects back through the output wire to MCU. Maybe some radiowave 
is
> also made, so I put the whole 220V contactors in a seperate metal 
case.
> 
> Question:
> 
> 1- What is the real frequency of this noise/spark/surge, and its
> voltage? If I know it, I can manage to attenuate it to tolerable 
level.
> 
> 2- How can I improve the situation? Maybe a low-pass filter do the 
job
> (or complete the isolation)?
> 
> 3- Is there a web link to information regarding contactor surge
> reduction? I couldn't find any. I probably can not find commercial
> surge killers, so please teach me how to make one with electronic 
parts.
> 
> [Non-text portions of this message have been removed]
> 
> 
> 
>                          
> 
>        
> ---------------------------------
> Never miss a thing.   Make Yahoo your homepage.
> 
> [Non-text portions of this message have been removed]
>